Corporate Affiliates Program
High Performance Computing Services
The Pittsburgh Supercomputing Center has been providing world-class computing power and balanced infrastructure for industry and leading academic institutions for more than two decades. The center provides an integrated array of services necessary for a company to approach problems previously viewed as beyond reach.
Among the HPCC products and services we offer are computing resources, expertise in high-performance networking, consulting and training.
Computing
- The latest high performance computing systems available at PSC bring our users unique capability in both speed and memory. The CRAY XT3, PSC’s most advanced leading-edige computing resource, is ideally suited for large-scale demanding projects in computational chemistry, material science, drug design and fluid mechanics. Pople, an SGI Altix 4700 shared-memory NUMA system, is well suited to applications that require moderate parallelism with fast access to very large memory. The availability of these systems provides our corporate partners with exceptional computational resources. It also provides them with the opportunity to leverage our experience in building and operating very large clusters to add advantage to their own cluster implementation.
- Virtually unlimited storage is available provided by a
combination tape-and-disk system, which has a total capacity of 3.2 Petabytes.
- High speed data links are provided between all of our
major computing platforms and
our dedicated
system for storing data generated
on our systems. We can also provide storage for your own internally
generated data. PSC is currently working with a small biomedical
startup firm to provide the mass storage capability required
for its daily operations and archival storage.
- The most popular software packages are available for computational chemistry, engineering, biomedical databases and sequence analysis, neural sciences and materials science. We can also acquire other software if requested and if versions are available for our computers.
Networking
- We work with corporate clients and researchers to facilitate new
technology from research, through
engineering and operations, to production use.
- The center's networking personnel is available to consult
on:
- design, engineering, testing, maintenance and general
support for the hardware associated with
production and experimental network infrastructure
- implementation, maintenance, tracking and monitoring
network and system security
- diagnosing and troubleshooting internal and external network connections
- design, engineering, testing, maintenance and general
support for the hardware associated with
production and experimental network infrastructure
Consulting & Training
- Individual one-on-one consulting is available to provide support for all phases of problem evolution from inception to final results. This could include software recommendations, formulation of the problem, optimization of single processor performance, as well as communications. PSC also provide logistics support to coordinate timing and acquisition of all resources required to run a job.
- Training in basic parallel programming techniques is available covering the entire spectrum of the problem solving procedure - from problem decomposition to message passing programming. This includes MPI and shared memory paradigms as well as optimization techniques. We cultivate codes that are portable and can grow with the research problems being tackled. Onsite and Colocation classes are provided to teach how to optimize codes for a specific type of system. Upon the request of corporate clients, we developed and taught a module on "Recommended Programming Practices," which demonstrates a standards-based, best-practices approach to software engineering for engineers and researchers.
Learn more about PSC's consulting services.